Article pubs.acs.org/ac
A Microfluidic SPLITT Device for Fractionating Low-Molecular Weight Samples Tristan F. Kinde and Debashis Dutta* Department of Chemistry, University of Wyoming, Laramie, Wyoming 82071, United States S Supporting Information *
ABSTRACT: In this article, we report the design of a microfluidic split flow thin cell (SPLITT) fractionation device with internal electrodes placed across the width of its analysis channel for assaying low-molecular weight samples. The reported device allows the realization of lateral electric fields and separation distances of the orders of 100 V/cm and 500 μm, respectively, that are suitable for fractionating such mixtures with high resolution. Our experiments show that a key challenge to realizing electrophoretic fractionations using the current design is to minimize the electroosmotically driven fluid circulations in its SPLITT channel that tend to hydrodynamically mix the liquid streams flowing through this duct. The present work addresses this challenge by chemically modifying the surface of our fluidic conduits with a new coating medium, N-(2-triethoxysilylpropyl) formamide, which has been shown to diminish electroosmotic flow in glass microchannels by over 5 orders of magnitude. Finally, we describe the integration of the reported microfluidic fractionation device to a mass spectrometer via the electrospray ionization interface to allow inline labelfree detection of analytes in our assay. Product purity greater than 95% has been accomplished using the SPLITT system presented here for a sample of peptides having the same electrical polarity.
E
from proteins to colloidal particles.17−20 This method, originally proposed by Giddings over 4 decades back,17 has been realized on the microfluidic platform only recently.14,21,22 In their miniaturization effort, Gale and co-workers developed several SPLITT devices in which the electric field was applied across the narrowest dimension of the analysis channel, i.e., its depth, that varied from ∼30 to 175 μm. While such a design was proven useful for sorting nanoparticles with large electrophoretic mobilities and small diffusion coefficients, its applicability to enriching low-molecular weight samples is not as promising. This is because fast diffusive mixing across 100 μm or smaller separation distances often results in low yields and purities for analytes with molecular weights up to a few kilodaltons. For example, a tightly focused stream of a kilodalton-sized peptide having a molecular diffusivity (D) of 5 × 10−6 cm2/s will increase its spatial variance by about (2Dt)1/2 ≈ 100 μm over a separation time (t) of 10 s. In order to then separate two of these streams with a resolution greater than 1, it becomes necessary to draw their mean positions apart by at least a distance of 4(2Dt)1/2 ≈ 400 μm.23 Of course, this analysis assumes that the initial width of the sample stream to be zero and there are no mechanisms, e.g., hydrodynamic forces, Joule heating, etc., other than diffusion at play
xtraction of target molecules from complex mixtures is an important scientific goal to enable the production of highquality chemicals and drugs.1,2 Such processes moreover can allow sample analysis with greater reliability by minimizing interference from other species in the matrix as is often desired in mass spectrometry based quantitation methods.3 Over the past decade or so, microfluidic approaches to extraction processes have emerged as a particularly powerful tool due to their ability to distinguish entities based on mechanisms that are not very effective at the macroscale, e.g., diffusion,4,5dielectrophoresis,6,7 magnetophoresis,8,9optical,10,11 and acoustic12 sorting, etc. In addition, the utility of microfluidic systems in scaling down traditional fractionation techniques like electrophoresis has permitted the realization of higher purities and yields at faster speeds than previously possible.13,14 Interestingly, such miniaturization, in general, has also led to lower fabrication and operational costs, greater portability, and reduced sample sizes for fractionation assays. While the latter benefits are natural consequences of scale reduction, the observed improvements in the separation performance of microfractionation devices are associated with the greater control over transport processes, larger surface area to volume ratios, and smaller Joule heating in microfluidic channels.15,16 Among the various fractionation methods developed to date, the electrical split flow thin cell fractionation (SPLITT) technique has been of significant interest to separation scientists due to the ease with which it can be implemented as well as its applicability to a wide variety of analytes ranging © 2013 American Chemical Society
Received: March 20, 2013 Accepted: July 10, 2013 Published: July 10, 2013 7167
dx.doi.org/10.1021/ac400843s | Anal. Chem. 2013, 85, 7167−7172
Analytical Chemistry
Article
Figure 1. (a) Reactions describing the synthesis of the N-(2-triethoxysilylpropyl) formamide species (top) and capping of the glass surface with the same molecule (bottom). (b) Fluorescence measured for the liquid aliquot collected at reservoir 2 of the device shown here after flowing rhodamine B through the microchannel fabricated on it both in the presence and absence of the N-(2-triethoxysilylpropyl) formamide coating. The “diffusion limit” here corresponds to the case when the neutral tracer was allowed to migrate from reservoir 1 to 2 of the same device by diffusion only.
■
EXPERIMENTAL PROCEDURE For fabricating the microfluidic devices employed in this work, bottom substrates and cover plates made from borosilicate glass were purchased from the Telic Company (Valencia, CA). The purchased cover plates and the bottom substrates came with a thin layer of chromium and photoresist laid down on one of their surfaces to enable the photopatterning process. Customdesigned photomasks created through Fineline Imaging Inc. (Colorado Springs, CO) were used to pattern the desired channel and electrode layout onto the bottom substrate and cover plate, respectively, using standard photolithographic methods.28 The width of the SPLITT channel varied between 300 and 900 μm in our device, while the electrode patterns in it were chosen to be either 0.5 mm or 1 mm wide. The channel segments leading to and exiting from the SPLITT conduit were exactly half as wide as the latter in all our designs. After completion of the photopatterning process, the photoresist layer was cured in microposit developer MF-319 (Rohm and Haas) and the chromium layer removed along the channel network with a chromium etchant (Transene Inc.). The fluidic ducts on the bottom substrate and electrode patterns on the cover plate were then etched to depths of 40 μm and 400 nm, respectively, using buffered oxide etchant (Transene Inc.). A 130 nm thick layer of chromium was later deposited within the electrode patterns followed by a 40 nm thick layer of gold using a dual metal evaporator system (Energy Beam Sciences, Inc.). The protective photoresist and chromium layers were subsequently removed from the bottom substrate and cover plate using the MF-319 and chromium etching solutions, respectively. Finally, the two glass plates were aligned and bonded using sodium silicate as an adhesive layer based on procedures described in the literature.29 A detailed schematic of the electrode fabrication process has been included in the Supporting Information. Access holes were drilled on the bottom substrate at appropriate locations prior to this bonding process, however, using a powder blaster system (Vaniman Manufacturing Co.) to enable electrical contact with the microfabricated electrodes as well as introduction of liquid streams into the fluidic network. The separation voltage across the SPLITT channel was applied by filling the relevant access holes with an electrically conductive paste and then immersing copper wires into it that were connected to the leads of an inhouse built 12 V dc power source.
contributing to the band broadening in the system. Keeping in mind that many of these assumptions do not hold true in practical situations and the mobility of the trailing analyte could be nonzero in the direction of migration of the leading one, it becomes often necessary to employ separation distances of the order of 500 μm or greater in order to fractionate low molecular weight mixtures with high resolution. One approach to realizing such separation distances (w) is to apply the electric field across the width of the SPLITT channel rather than its depth. Because the time scales for electrophoresis and diffusive mixing vary with w and w2, respectively, the former process tends to dominate the lateral migration of analyte molecules in this design.23 Unfortunately, the application of an electric field across the width of a SPLITT channel also results in transverse electroosmotic flow in it creating fluid circulations.24 Such circulations have the tendency to vigorously mix the liquid streams flowing through the channel compromising the resolution of the desired fractionations. In this article, we demonstrate that in order to continually fractionate lowmolecular weight mixtures with high resolution through application of an electric field across the width of a microfluidic SPLITT channel, it is necessary to reduce the electroosmotic flow (EOF) in the system by several orders of magnitude. While static coatings commonly used in capillary electrophoretic systems, e.g., polyacrylamide, can diminish EOF in glass capillaries by a factor of 10 or so,25 they are not as effective in preventing hydrodynamic mixing of flow streams in the fractionation device described above. Here we report a new coating medium, N-(2-triethoxysilylpropyl) formamide,26,27 that has been shownto diminish EOF in glass microchannels by over 5 orders of magnitude, practically eliminating the unwanted mixing observed in a microfluidic SPLITT device with electrodes placed across the width of its analysis channel. This minimization of fluid mixing then enables high-resolution fractionation of low-molecular weight samples in our system which can be quantitated continually with a mass spectrometer via the electrospray ionization interface. Product purities greater than 95% have been accomplished using the reported microfluidic SPLITT device for separations involving dye and peptide mixtures. Moreover, the current system has been shown to be capable of monitoring near real time changes in the composition of the sample stream with a response time of about 25 s. 7168
dx.doi.org/10.1021/ac400843s | Anal. Chem. 2013, 85, 7167−7172
Analytical Chemistry
■
The reported microchip was prepared for an experiment by treating the entire fluidic network with 1 M NaOH for an hour followed by sequentially rinsing with water, methanol, and acetone for 10 min each. The microchannels were then derivatized with a solution of N-(2-triethoxysilylpropyl) formamide for 2 h at room temperature to suppress electroosmotic flow in the system.26,27 This coating material was prepared by mixing stoichiometric amounts of commercially available 3-aminopropyltriethoxysilane with ethyl formate and reacting the two chemicals under ambient conditions for 48 h (reaction shown in Figure 1a). The sample used in this work was either a solution of 100 μM rhodamine B, a mixture of 100 μM each of rhodamine 6G, and fluorescein or a solution containing the peptides MRFA (molecular weight (mol. wt.) 523.1 Da) and γ-2-MSH (mol. wt. ∼1569.4 Da) at concentrations of 14 μg/mL and 48 μg/mL, respectively, as was relevant to the experiment. All of these samples were prepared in deionized water containing 25% methanol. Syringe pumps were used to introduce the solvent and sample streams into our SPLITT device through ports 2 and 1, respectively. In the current study, these streams were always pumped at the same flow rate for the dye-based experiments but at a ratio of 10:1 (solvent/sample) in the peptide-based ones. While the composition of the stream leaving the SPLITT channel was determined by directly electrospraying this liquid into a massspectrometer for the peptide sample, the same task was performed for the dye mixture by flowing about 100 μL of this liquid into reservoirs placed at the downstream end of our fluidic network. In the latter case, 20 μL aliquots of the liquid collected in these reservoirs were subsequently pipetted out and diluted to 4.41 mL in a standard 1 cm × 1 cm quartz cuvet, prior to measuring their fluorescence using a commercial spectrophotometer. The emission intensities for rhodamine B, rhodamine 6G, and fluorescein were measured at 574, 547, and 511 nm, respectively, in this study for a common excitation wavelength of 490 nm. Electroosmotic flow measurements in glass microchannels were performed in our current work using the microchip device shown in Figure 1b that comprised a single 3 cm long, 2 mm wide, and 40 μm deep straight conduit. Notice that we could not have performed these measurements by injecting a neutral tracer into the analysis duct of a cross-channel layout,30−32 because the EOF in our N-(2-triethoxysilylpropyl) formamide coated conduits was observed to be too weak to even realize sample injection. After preparing this microchip following the procedures described in the previous paragraph, a 100 μL sized 1 mM rhodamine B (neutral tracer) sample prepared in deionized water containing 25% methanol was placed in reservoir 1, while filling its reservoir 2 with the dye free solvent. A high voltage (1 kV) was subsequently applied to the sample reservoir for a 30-min period following which a 20 μL aliquot of liquid was collected from reservoir 2. While this aliquot was diluted to 4.41 mL (a dilution factor of 220.5) prior to measuring its fluorescence for estimating the EOF rate in the coated microchip, the dilution factor chosen for the uncoated channel case was 5000 times larger, i.e., 5000 × 220.5, due to the large EOF observed in that situation. To minimize any flow of the dye due to a difference in the hydrostatic heads at the channel terminals, fluid reservoirs with a large diameter (3 cm) were affixed to these ends and the liquid levels in them were equated at the start of all of our experiments.
Article
RESULTS AND DISCUSSION
Suppression of Electroosmotic Flow. In Figure 1b, we have presented the fluorescence profiles for the solutions collected from reservoir 2 of the single-channel device, which translate to time-averaged rhodamine B fluxes of 3 × 10−3 and 2.7 × 10−8 mol/m2/s through the uncoated and coated microchips, respectively. This corresponds to a reduction in the electroosmotic flow in our glass microchannel by over 5 orders of magnitude using the N-(2-triethoxysilylpropyl) formamide coating. In addition, we also performed a control experiment in which the neutral tracer was allowed to migrate from reservoir 1 to 2 by diffusion only. This experiment yielded a timeaveraged flux value (3.6 × 10−10 mol/m2/s) about 75 times smaller than that observed in the coated microchannel case under electroosmotic flow conditions. Notice that even with a large EOF, the rise in liquid level in reservoir 2 of our uncoated device was less than 0.5 mm after the 30-min experimental period. On the basis of the Poiseuille equation for a parallelplate geometry,33−36 this hydrostatic head was estimated to yield a pressure-driven backflow that was over 100-fold weaker than the electroosmotic transport rate in a bare glass channel. Consequently, the error in our EOF measurement arising from this backflow between reservoirs was expected to be less than a percent of the value reported here. In addition to quantitating the ability of N-(2-triethoxysilylpropyl) formamide to suppress EOF in glass microchannels, we also investigated the stability of this coating material in our experimental system. It was found that the EOF rate in a glass conduit derivatized with this chemical varied by less than 5% over a 3-week period when the coated duct was rinsed with methanol for 10 min after use, followed by drying it in an oven for an hour at 80 °C before storing it at 4 °C. Establishment of Fractionation Conditions. The utility of the N-(2-triethoxysilylpropyl) formamide coating in minimizing lateral mixing of flow streams in the SPLITT device depicted in Figure 2a was initially evaluated by flowing the solvent and rhodamine B sample through it at 3 μL/min
Figure 2. (a) The top image here depicts the SPLITT device used in this work. The bottom fluorescence images here show the distribution of the neutral dye, rhodamine B, across the width of our fractionation channel 1.5 cm downstream from its entrance in the presence of a lateral electric field. (b) Ratio of the full width at half-maximum for the fluorescence profile in the analysis duct to the channel width, i.e., mixing parameter (θ), measured at different locations along the length of the SPLITT conduit. A 100 μM sample of rhodamine B prepared in deionized water containing 25% methanol (by volume) was used in this study. The flow rates for the dye and solvent streams in the experiments depicted in parts a and b were each set to 3 μL/min, with 5 V applied across the microfabricated electrodes. 7169
dx.doi.org/10.1021/ac400843s | Anal. Chem. 2013, 85, 7167−7172
Analytical Chemistry
Article
Figure 3. (a) Fluorescence measured for the liquid aliquot collected at reservoirs 3 and 4 of our SPLITT device after electrophoretically extracting the cationic dye from the rhodamine 6G and fluorescein mixture for 30 min. In this experiment, the solvent and dye streams were transported at 3 μL/min each, applying 5 V across the width of the analysis channel. (b) Ratio of the molar concentrations of rhodamine 6G to fluorescein (α) as a function of the sample flow rate through our SPLITT device. The flow rates for the solvent and dye streams were always chosen to be equal to each other in this study. All fractionations shown in parts a and b were run for 30 min, and the fluorescence readings for the liquid aliquots collected from the downstream reservoirs were made after diluting them by a factor of 220.5.
appropriately diluted liquid aliquots collected from the downstream reservoirs of this device after the sample was streamed at a rate of 3 μL/min through a 600 μm wide analysis channel applying a separation voltage (φ) of 5 V for 30 min. Quantitation of these peaks shows (calibration data not included) that the concentrations of rhodamine 6G and fluorescein in reservoir 3 to be 96.2 μM and 3.2 μM, while that for the same dyes in reservoir 4 to be 3.6 μM and 97.1 μM. These numbers correspond to purities of 96.8% and 96.4% for rhodamine 6G and fluorescein in reservoirs 3 and 4, respectively, closing the mass balance for these dyes to within 1%. As part of this study, we also looked into the effect of channel width on the performance of our SPLITT device in an attempt to achieve the highest purity of the cationic dye. Results obtained from these experiments have been presented in Figure 3b in which the ratio of molar concentrations for rhodamine 6G to fluorescein (α) in reservoir 3 at the end of the 30-min fractionation period has been used as the figure-ofmerit. The figure shows that for any chosen fractionation voltage and channel width, the value of α is reduced somewhat with an increase in the sample flow rate due to a decrease in the separation time. At a fixed sample flow rate and channel width, however, there exists an optimum separation voltage at which the purity of the cationic dye in reservoir 3 is maximized. This maximum occurs at the highest voltage (φmax) that can be applied across the microfabricated electrodes without any observed electrochemical gas generation in the system. While the product purity in reservoir 3 decreases gradually with a reductionin φ for φ < φmax due to a weakening separation field (see data presented for the 900 μm wide channel), this quantity is seen to rapidly approach a value of 1 with an increase in φ for φ > φmax as a result of fluid mixing introduced by the electrochemically generated gas bubbles (see data presented for the 300 μm wide channel). Finally, it also turns out that when operating at φmax the reported SPLITT device yields the best result for the 600 μm wide conduit among the 3 channel widths tested here using the rhodamine 6G/fluorescein sample. Notice that the smaller values of α yielded by the 900 μm wide channel here results from a weak separation field (at φmax) and a long lateral migration distance compared to the other two cases. On the other hand, faster diffusion of dye molecules across
each. Fluorescence images were taken under steady-state conditions at different locations of the analysis channel to then visualize the lateral mixing process both in the presence and absence of a transverse electric field. In this study, 5 V was applied across a 600 μm wide SPLITT conduit which was determined to be the maximum applicable voltage under the experimental conditions before electrochemical gas generation at the electrodes started affecting the fluid flow profile in the system. Two representative images from these experiments have been included in Figure 2a that depict the extent of hydrodynamic mixing of the solvent and sample streams after flowing a distance of 1.5 cm in the analysis channel. The images show that while these liquid streams are rapidly mixed in an uncoated SPLITT channel by the electroosmotically driven fluid circulations, such mixing is practically eliminated in a conduit coated with N-(2-triethoxysilylpropyl) formamide. We have further examined these experimental images by quantitating the fluorescence profile recorded in them, and expressing the spatial homogeneity in these profiles across the channel width in terms of a mixing parameter (θ). This parameter, estimated as the ratio of the full width at half-maximum for the fluorescence profile to the channel width, has been plotted in Figure 2b as a function of the distance traveled by the solvent and sample streams together. An in-house written imageanalysis code based on MATLAB was used to estimate the values of θ reported here. The measurements presented in Figure 2b verify that the rate of migration of rhodamine B molecules into the solvent stream in our N-(2-triethoxysilylpropyl) formamide coated SPLITT channel indeed is dominated by diffusion and practically unaffected by the presence of a lateral electric field. Moreover, it shows that a polyacrylamide coating, commonly employed in capillary electrophoretic systems for reducing EOF, is not as effective in accomplishing the same goal in our SPLITT device. The polyacrylamide species was applied to the surface of our analysis channel in this study following procedures previously described in the literature.37 Proceeding further, the reported SPLITT microdevice was applied to fractionating a mixture of two fluorescent dyes, rhodamine 6G (cationic) and fluorescein (anionic). In Figure 3, we have presented the fluorescence spectrum for the 7170
dx.doi.org/10.1021/ac400843s | Anal. Chem. 2013, 85, 7167−7172
Analytical Chemistry
Article
streamlines tends to deteriorate the purity of rhodamine 6G in reservoir 3 of our SPLITT device with a 300 μm wide channel. The yield and purity of the cationic dye was observed to exceed 96% in the 600 μm wide analysis duct for sample flow rates smaller than 3 μL/min and a separation voltage of 5 V (φmax). Label-Free Fractionation of a Peptide Mixture. We finally assessed the ability of our SPLITT system to assay complex mixtures by applying it to the fractionation of two peptides, MRFA and γ-2-MSH, having the same electrical polarity. Notice that fractionations of this kind are particularly difficult to accomplish in a microfluidic SPLITT device as fast diffusion tends to dominate the lateral migration of small molecules in these systems. In order to further broaden the utility of our microfractionation unit, we also integrated it to a mass spectrometer (MS) via the electrospray ionization (ESI) interface to allow inline label-free detection of analytes. Interestingly, the solvent used in the dye-separations previously, i.e., 25% methanol (v/v) in deionized water, turns out to be an MS-friendly one and was therefore selected for the current assay as well. The concentrations for the peptide species in our sample were chosen such that they yielded about the same signal (relative abundance) when introduced into the mass spectrometer individually. The MS instrument was interfaced to our microchip by connecting a commercial ESI nozzle (New Objective, Inc.) to port 3 through a 25 cm long-75 μm i.d. glass capillary and an appropriate flow adaptor (IDEX Corporation). This integration however increased the hydrodynamic resistance of the flow path to the electrospray interface, which in turn led to an unequal splitting of the fluid stream exiting the separation channel toward ports 3 and 4. We rectified this issue by appending a flow adaptor-capillary-ESI nozzle assembly to port 4 as well, identical to that used in interfacing our microchip to the MS. With the fluid stream now splitting equally toward ports 3 and 4 and the analytes electrophoretically migrating in the same direction, it was realized that in order to accomplish the desired fractionation, the sample stream needs to be pinched to a small fraction of the channel width at the entrance of the analysis duct. To this end, the solvent flow into our device was chosen to be 10 times greater than that of the sample. The separation time in these experiments was unaltered from the previous fractionation assay, i.e., 8.4 s, by maintaining an overall liquid flow of 6 μL/ min through our 600 μm wide and 40 μm deep SPLITT channel. The current peptide fractionation was quantitated in terms of the relative abundance signal measured by the MS instrument as a function of the separation voltage (φ). It must be noted that after each increment in the value of φ, the device was run for 2 min to allow it to reach a steady-state. In Figure 4, we have presented this steady-state signal which shows the lighter peptide (MRFA) in our sample to begin appearing in the liquid stream reaching the MS instrument when φ = 2.6 V. With an increase in the separation voltage beyond this value, the MRFA concentration in the flow stream reaching port 3 rises sharply attaining its saturation value at φ = 3.2 V. Interestingly, the heavier peptide in our sample (γ-2-MSH) starts emerging in the mass spectrum right around this separation voltage and attains its saturation level at φ = 4.4 V. If we assume that the relative abundance recorded by the MS instrument to be proportional to the concentration of peptides in the electrosprayed liquid, Figure 5a shows that our SPLITT device is able to enrich the MRFA species in the flow stream reaching port 3 by as much as a factor of 21. This enrichment corresponds to over 95% purity of the lighter peptide realized
Figure 4. Electrophoretic fractionation of two peptides (MRFA and γ2-MSH) having the same electrical polarity using our SPLITT device integrated to a mass spectrometer via the electrospray ionization interface. In this experiment, the flow rate for the solvent stream was chosen to be 10 times greater than that for the sample. Included in the figure are also the recorded steady-state mass spectra at separation voltages 3.2 and 5.0 V. The relative abundances reported for MRFA and γ-2-MSH in this figure correspond to the height of the peak at m/ z of 524.1 and 1570.4, respectively.
at an optimum separation voltage of 3 V. As part of this study we also looked into the dynamics of the signal recorded by the MS instrument when the separation voltage undergoes a step change. Figure 5b shows the temporal variation in the measured relative abundance for the MRFA species upon increasing the value of φ from 2.6 V to 3.0 V in increments of 0.2 V. As may be seen from the figure, there is a delay of about 15 s from the instant the separation voltage is changed to the point the MS signal starts increasing. This delay corresponds to the dead volume between the microchip and mass spectrometer included in the flow adaptor-capillary-ESI nozzle assembly. Following this delay, the relative abundance is seen to increase sharply to its next steady-state value in about 10 s, which relates to the transit time of the peptides in the SPLITT channel. Overall, the response time for our fractionation system is about 25 s, which offers the potential for monitoring near real-time changes in the sample stream composition. Of course, this response time may be reduced by as much as a factor of 2 by simply decreasing the dead volume in the flow adaptor-capillary-ESI nozzle assembly over that reported here.
■
CONCLUSIONS To summarize, we have successfully demonstrated the operation of a microfluidic electrical fractionation system that is suitable for analyzing low-molecular weight mixtures. The reported fractionation was accomplished in a 40 μm deep and 600 μm wide channel with microfabricated electrodes placed across its width. Lateral hydrodynamic mixing was practically eliminated in this design through the use of a novel N-(2triethoxysilylpropyl) formamide coating that was shown to reduce electroosmotic flow in glass microchannels by over 5 orders of magnitude. The reported microchip was further integrated to a mass spectrometer via the electrospray ionization interface to allow inline label-free quantitation of the fractionated analytes. The functionality of our SPLITT unit was finally established by analyzing a mixture of two peptides, MRFA and γ-2-MSH, having the same electrical polarity and enriching the concentration of the former species by a factor of 21 in the product stream under optimum conditions. Interestingly, the current design also yielded a response time 7171
dx.doi.org/10.1021/ac400843s | Anal. Chem. 2013, 85, 7167−7172
Analytical Chemistry
Article
Figure 5. (a) Ratio of the measured relative abundances for MRFA to that for γ-2-MSH as a function of the separation voltage in our SPLITT device. (b) Temporal variation in the measured mass spectrometer signal (relative abundance) for the MRFA species when the separation voltage in our SPLITT device underwent a step change. The relative abundances reported for MRFA and γ-2-MSH in this figure correspond to the height of the peak at m/z of 524.1 and 1570.4, respectively. (11) Lee, K. H.; Kim, S. B.; Lee, K. S.; Sung, H. J. Lab Chip 2011, 11, 354−357. (12) Laurell, T.; Petersson, F.; Nilsson, A. Chem. Soc. Rev. 2007, 36, 492−506. (13) Pamme, N. Lab Chip 2007, 7, 1644−1659. (14) Narayanan, N.; Saldanha, A.; Gale, B. K. Lab Chip 2006, 6, 105−114. (15) Gale, B. K.; Caldwell, K. D.; Frazier, A. B. Anal. Chem. 2001, 73, 2345−2352. (16) Sant, H. J.; Gale, B. K. J. Chromatogr., A 2006, 1104, 282−290. (17) Caldwell, K. D.; Giddings, J. C.; Myers, M. N.; Kesner, L. F. Science 1972, 176, 296−298. (18) Tri, N.; Caldwell, K.; Beckett, R. Anal. Chem. 2000, 72, 1823− 1829. (19) Sant, H. J.; Chakvarty, S.; Merugu, S.; Ferguson, C. G.; Gale, B. K. Anal. Chem. 2012, 84, 8323−8329. (20) Somchue, W.; Siripinyanond, A.; Gale, B. K. Anal. Chem. 2012, 84, 4993−4998. (21) Gale, B. K.; Caldwell, K. D.; Frazier, A. B. Anal. Chem. 2002, 74, 1024−1030. (22) Fennah, M.; Manz, A. In Proceedings of the μTAS 2002 Symposium; Kluwer Academic Publishers: Dordrecht, The Netherlands, 2002, Vol. 2, pp 817−819. (23) Jorgenson, J. W.; Lukacs, K. D. Anal. Chem. 1981, 53, 1298− 1302. (24) Lynn, N. S.; Henry, C. S.; Dandy, D. S. Microfluid. Nanofluid. 2008, 5, 493−505. (25) Doherty, E. A. S.; Meagher, R. J.; Albarghouthi, M. N.; Barron, A. E. Electrophoresis 2003, 24, 34−54. (26) Toh, G. M.; Yanagisawa, N.; Corcoran, R. C.; Dutta, D. Microfluid. Nanofluid. 2010, 9, 1135−1141. (27) Toh, G. M.; Corcoran, R. C.; Dutta, D. J. Chromatogr., A 2010, 1217, 5004−5011. (28) Reyes, D. R.; Iossifidis, D.; Auroux, P. A.; Manz, A. Anal. Chem. 2002, 74, 2623−2636. (29) Wang, H. Y.; Foote, R. S.; Jacobson, S. C.; Schneibel, J. H.; Ramsey, J. M. Sens. Actuators, B 1997, 45, 199−207. (30) Yanagisawa, N.; Dutta, D. Electrophoresis 2010, 31, 2080−2088. (31) Xia, L.; Dutta, D. Anal. Chem. 2012, 84, 10058−10063. (32) Xia, L.; Dutta, D. Analyst 2013, 138, 2126−2133. (33) Bird, R. B.; Stewart, W. E.; Lightfoot, E. N. Transport Phenomena; John Wiley & Sons: Singapore, 1994. (34) Dutta, D. Microfluid. Nanofluid. 2011, 10, 691−696. (35) Dutta, D.; Ramsey, J. M. Lab Chip 2011, 11, 3081−3088. (36) Dutta, D. Chem. Eng. Sci. 2013, 93, 124−130. (37) Ramsey, R. S.; Ramsey, J. M. Anal. Chem. 1997, 69, 1174−1178.
of about 25 s offering the potential for monitoring near realtime changes in the sample stream composition.
■
ASSOCIATED CONTENT
S Supporting Information *
Detailed schematic of the electrode fabrication process. This material is available free of charge via the Internet at http:// pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This research work was supported by the Defense Threat Reduction Agency (DTRA) through contract HDTRA1-09-C0013. D.D. also acknowledges funds from the National Science Foundation and Wyoming INBRE program through Grants CBET−0854179 and P20RR016474, respectively, for completing some of the experiments included in this manuscript. The authors thank Prof. Robert C. Corcoran for assisting with the development of the N-(2-triethoxysilylpropyl) formamide coating material.
■
REFERENCES
(1) Lord, H.; Pawliszyn, J. J. Chromatogr. 2000, 902, 17−63. (2) Lee, J. Y.; Lee, H. K.; Rasmussen, K. E.; Pedersen-Bjergaard, S. Anal. Chim. Acta 2008, 624, 253−268. (3) Antignac, J. P.; de Wasch, K.; Monteau, F.; De Brabander, H.; Andre, F.; Le Bizec, B. Anal. Chim. Acta 2005, 529, 129−136. (4) Brody, J. P.; Osborn, T. D.; Foster, F. K.; Yager, P. Sens. Actuators, A 1996, 54, 704−708. (5) Brody, J. P.; Yager, P. Sens. Actuators, A 1997, 58, 13−18. (6) Hughes, M. P. Electrophoresis 2002, 23, 2569−2582. (7) Li, Y. L.; Dalton, C.; Crabtree, H. J.; Nilsson, G.; Kaler, K. V. I. S. Lab Chip 2007, 7, 239−248. (8) Pamme, N. Lab Chip 2006, 6, 24−38. (9) Kang, J. H.; Park, J. K. Small 2007, 3, 1784−1791. (10) MacDonald, M. P.; Spalding, G. C.; Dholakia, K. Nature 2003, 426, 421−424. 7172
dx.doi.org/10.1021/ac400843s | Anal. Chem. 2013, 85, 7167−7172
